Columbia Plain sherds

Figure I. Columbia Plain sherds (left to right) (top) SA2 from Convento de San Francisco, Dominican Republic ana SA41 from Juandolio, Dominican Republic (bottom) SA19 from Nueva Cadiz, Venezuela and SA59 from LdVeija, Dominican Republic...

Table III presents the proposed location of production on the basis of the chemical composition for the majolica types analyzed and the attributions made by Goggin (I) and Lister and Lister (2). All Mexico City White and Mexico City copies of Columbia Plain belong to the proposed Mexico City production as just discussed. All the San Luis Blue-on-White sherds reported here are from the excavations at Santa Catalina de Guale. With the exception of the three Spanish composition sherds discussed earlier, the remaining 17 sherds can be chemically classified as having been produced in Puebla. Both Goggin and the Listers propose a Mexico City production for San Luis Blue-on-White. Goggin s attribution to Mexico City is restricted to the reddish paste variant. We have no chemical confirmation of a Mexico City production for San Luis Blue-on-White.

Using the data in Table I for 40 Columbia Plain and Yaval-Blue-on-White sherds from Convento de San Francisco, Dominican Republic and for 41 sherds from Nueva Cadiz, Venezuela, we have calculated a value of t by dividing the diflEerence between the means of the concentrations by the standard error of the difference between the means. We have calculated t for the oxides of rubidium, sodium, and cesium, and the numbers obtained are 5.6, 4.1, and 1.8, respectively. With this sample size the significance levels for the values of t for rubidium and sodium are greater than 99.5% and for cesium are greater than 90%. There is, therefore, a statistically significant difference between the concentrations of each of these three oxides for all of the sherds from Convento de San Francisco compared with those from Nueva Cadiz. 

---